Periodontitis has been associated with systemic diseases such as cardiovascular disease and low birth weight. Establishing the causality of these associations is difficult; the compared groups, individuals with and without periodontitis, may differ on factors other than periodontitis. If these other factors are causally related to systemic disease risk, confounding may be present. The primary goal of this R03 proposal is to provide a methodological study of confounding by means of different approaches. This methodological work will be performed in three large databases: the First National Health & Nutrition Examination Survey (NHANES I) (n=31,973), the NHANES I Epidemiological Follow-up Study (NHEFS) (n=11,348), and the Third National Health & Nutrition Examination Survey (NHANES III) (n=30,818). Exhaustive information regarding most potentially confounding variables is present in these datasets (greater than 1000 variables per individual). Preliminary analyses of the NHANES I and NHEFS suggest that confounding is a significant methodological challenge. Periodontitis was associated with unexpected differences that may in and of itself cause other systemic diseases. Individuals with periodontitis, after adjustment for age and gender, were more likely to have an unhealthy diet (less fruit, less vitamin C, less riboflavin, and less linoleic acid), were more likely to have suffered a nervous breakdown, less likely to exercise, more likely to be overweight, and more likely to have high blood pressure. In addition, after adjustment for known risk factors, periodontitis was associated with a two-fold increase in lung cancer mortality, an unexpected finding that suggests that confounding may have a strong influence. The aims of this study are: 1) to provide a systematic evaluation of potentially confounding variables in NHANES I and NHANES III; 2) to provide a systematic evaluation of the association between periodontitis and the most common causes of mortality in NHEFS, 3) to assess the benefits of using alternate referent or unexposed groups in cohort studies in NHEFS; and 4) to perform a series of sensitivity analyses to assess the extent to which confounding variables may cause spurious associations. The investigators believe the completion of these projects may play a pivotal role in determining whether the previously observed associations between periodontitis and systemic diseases are spurious or causal.